TEMPERATURE COEFFICIENT OF A HETEROZYGOTE. 109 



In this way the coefficient of dominance of full eye is readily deter- 

 mined as 5.9 per cent, at 15 , 4.3 per cent, at 20 , 2.4 per cent, at 

 2j° , and 2.9 per cent, at 30 °. The general decrease with increase 

 in temperature is evident. The reversal between 27 and 30 may 

 probably be explained by the disturbance resulting from an ap- 

 proach to the maximum temperature. 



The Value of a Germinal Factor in Terms of an Environmental 

 One. — Perhaps the most interesting point in connection with the 

 present data is the demonstration that they furnish of the fact that 

 the gene, ultra-bar, has the same type of reaction as a temperature 

 difference. It is possible to state the effectiveness of particular 

 germinal factors in terms of the corresponding effects of tempera- 

 ture. Such an attempt has been made in previous studies of the 

 bar races and the temperature coefficient has given the basis for 

 the evaluation. Since these previous studies had shown that a 

 change of one degree in temperature produces a change of approxi- 

 mately 10 per cent, in ommatidial number, a " unit " factor was, for 

 convenience, taken as one that produces the same change. The 

 units in the factorial scale, whether environmental or germinal, are 

 thus expressed on the same basis. 



First of all, it will be well to take up the demonstration of the 

 fact that the particular germinal difference represented by the 

 addition of a second ultra-bar factor in place of the full factor of 

 the heterozygote does not correspond to a constant somatic expres- 

 sion. The difference between heterozygous and homozygous ultra- 

 bar is not represented by a constant difference in ommatidial num- 

 ber. Thus at 15 degrees the difference is 60.6 ommatidia, at 20 

 degrees 39.5, at 27 degrees 19.2, and at 30 degrees 21.3. These 

 data are given in Table I. and in graphic form in Fig. 2, where the 

 lengths of the heavy vertical lines are proportional to the omma- 

 tidial differences at the various temperatures. The marked change 

 with temperature is obvious, though the germinal difference re- 

 mains constant. The ommatidial difference can not, therefore, 

 serve directly as a measure of germinal difference. 



If, however, the unit of measurement is the effect produced by 

 a degree of change in temperature, 8 per cent, in this case, and the 

 effect produced by the substitution of the second ultra-bar factor 

 for the full factor of the heterozygote is measured in terms of this 



